lmb-freiburg/hand3d
A convolutional neural network that estimates 3D hand pose from a single RGB image using deep learning.
Not currently ranked — collecting fresh signals.
star history
ColorHandPose3D is a deep learning system that predicts 3D hand pose from single color images using a convolutional neural network. The network was developed at the University of Freiburg and published at ICCV. It includes training and evaluation scripts with preprocessing pipelines for standard hand pose datasets (RHD and STB). The implementation uses TensorFlow and ships with pre-trained model weights and example data for forward-pass inference.
Frequently asked
- What is lmb-freiburg/hand3d?
- A convolutional neural network that estimates 3D hand pose from a single RGB image using deep learning.
- Is hand3d open source?
- Yes — lmb-freiburg/hand3d is open source, released under the GPL-2.0 license.
- What language is hand3d written in?
- lmb-freiburg/hand3d is primarily written in Python.
- How popular is hand3d?
- lmb-freiburg/hand3d has 812 stars on GitHub.
- Where can I find hand3d?
- lmb-freiburg/hand3d is on GitHub at https://github.com/lmb-freiburg/hand3d.